February 01, 2015, 02:42:52 PM

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - dtaylor

Pages: [1] 2 3 ... 57
EOS Bodies / Re: Canon 7Dii vs Samsung NX-1 Epic Shootout
« on: January 19, 2015, 09:21:14 PM »
Nice review.  I was disappointed to see the Samsung, even with its BSI sensor, do worse at high ISO noise even compared to Canon's crop sensor.

You say that as if Canon's crop sensor is poor at high ISO. The 7D2 is clearly at the top of the APS-C list right now, the 70D is quite good, and even the current 18 MP cameras aren't bad.

Lenses / Re: Canon EF 400mm f/4 DO IS II USM Lens Review
« on: January 18, 2015, 09:26:51 AM »
The MTF Chart in this review indicates that the lens ist still "excellent" at f8 and f11 but I still find that odd. Often people stop down to f8 or f11 to increase depth of field so if sharpness decreases after f5.6 ...

If the lens is diffraction limited then the sharpness at those apertures would be no worse then any other lens, and quite likely better.

A lens that improves as it's stopped down a bit is fighting diffraction and lens aberrations. The best it can do is hit an aperture setting where the aberrations become insignificant and the sharpness/resolution is bound by diffraction...like the 400 DO II appears to be at any aperture.

Put another way you're not losing anything at f/8 or f/11...you are gaining something at f/4.

Lenses / Re: Why not uppdate just the IS
« on: January 17, 2015, 11:36:44 PM »

Based on what? The old adage that primes are always better then zooms hasn't been true in years thanks to modern lens design and manufacturing. The 100-400 II was a significant update and no primes blow it away now.

Are you sure?

Quite sure on both statements.

Yes, some of the new zooms are better than the old primes, but given the same technology and materials the prime will always be better than the zoom.

A prime cannot be better then a diffraction limited zoom. And yes, someday soon we will start to see diffraction limited primes and zooms for small format. (A few such lenses exist for medium format.)

We're not at that point yet in small format lenses. But as we approach that point we see diminishing returns in IQ gains. Improved tech benefits the zooms more then the primes and closes the gap. This is exactly what we're seeing as Canon (and others) update their lenses.

The prime will have less elements, and that means less transmissive loss and less surfaces for reflection.

Given modern coatings this is rarely, if ever, human observable.

The prime will not have to worry about alignment of the zoom mechanism.

Given the complexity of any lens design which approaches diffraction limited IQ this is a non-issue. Either will have very complex alignment and manufacturing tolerances. I don't think there's a white tele shipping today with less then a dozen elements.

The prime will be mechanically superior.

How so?

Look at the two MTF charts.... which one is for a brand new "state of the art" zoom, and which one is for a couple of years old prime?

Did you bother to look at test shots for each? The differences are smaller then your post processing choices. While the prime is better, it does not by any stretch of the imagination "blow away" the zoom.

The prime is amazing in that there's not much IQ difference between f/2.8 and f/5.6 where the zoom starts at f/5.6. But I highly doubt you could tell unlabeled prints from these lenses apart.

Some background: with small format if I can clearly see a difference in unlabeled 24" prints I consider it significant. If I can only see it in 36" prints then I consider it minor. If I can't reliably detect the difference in unlabeled 36" prints then I consider it "tiny" and inconsequential no matter what it looks like on a graph or when pixel peeping B&W line charts.

"Unlabeled" is key because human psychology being what it is we will perceive one to be better if we expect it to be. I don't mean to say that someone will lie because they want their lens/camera/wine/whatever to win. Your personal experience of A will literally be better if you expect, or are told, that A is better then B before the test. Take away the labels and the results are often surprising.

I believe you would be very hard pressed to detect any of the lenses we're discussing given 36" prints of real world scenes, especially if they have been processed.

Lenses / Re: Why not uppdate just the IS
« on: January 17, 2015, 11:07:06 PM »
I also took stock of DTP to gauge the sharpness and you seem to agree that the 300L IS looks no sharper than the new 100-400L IS. To me that is unconvincing performance.

Then the 100-400 mark II, yes. It's quite a bit sharper then the mark I.

What was "excellent" prime performance yesterday is not "excellent" when new tech produces ever better lenses.

The mark II isn't better, though it pulls even. You're assuming that for a given level of lens design/manufacturing tech that primes should always be much better then zooms. That's a false assumption.

None of Canon's primes, not even their most expensive and newest mark II super telephotos, are significantly sharper then the 300 f/4L IS OR the 100-400 mark II. Their most recent L zooms...the 70-300L, 100-400 II, and the latest 70-200L variants...quite dramatically closed the gap on all of their primes vs. the situation we had just a decade ago.

The same is true even for the Sigma ART and Zeiss Otus primes. Sharper then the L zooms? Yes. But by small amounts.

The way you described the 300 f/4L IS someone might think it's mediocre and that a new version would be dramatically better. Old design or not, it's one of the sharper lenses in existence, and you have to look at the very best primes made to find even a slight improvement. At the level we're debating IQ gains are very small and very, very expensive.

That does not mean your 300IS L takes any lesser pictures than yesterday. But it shows that an upgrade could be even better. Why be troubled by that?

The gain seen in the 100-400 II does not mean an updated 300 f/4L IS would be much better. A perfect lens is diffraction limited wide open. If you look at the 300 f/2.8L II it's roughly one stop away from that right now. f/4 is better then f/2.8 (barely). There's no gain at f/5.6, and f/8 sees degradation. I don't expect a new 300 f/4L would do any better then this, and this is a relatively small gain in IQ vs. the current f/4L IS.

2) 70-300L IS II (which I also have) is only OK and cannot compare with the 300mm L IS

"Only OK?"  ::)

The edges are better on the primes, with mid-frame and center being only very slightly better (f/4L or f/2.8L II). But the high contrast TDP charts emphasize any differences. Given a real scene...good post processing technique...you would be hard pressed to pick unlabeled prints from one or the other.

Doesn't mean I wouldn't pick one of the primes. Aperture is aperture and even a single stop can be huge. But we are splitting hairs when debating the IQ of any of these lenses. EF-S 55-250 STM vs. one of the 300L primes...THERE'S a big difference, easily print visible, no post processing away that. 70-300L vs. one of the 300 primes? Eh...the Ferrari does 185 mph, the Porsche does 187 mph, either one will get you a nasty ticket  ;)

Lenses / Re: Why not uppdate just the IS
« on: January 17, 2015, 10:39:52 AM »
From what I have seen it has a hard time matching the new 100-400mm IS L II for sharpness.

What have you seen? Were they testing a flawed copy?

Head over to TDP. At 300mm they are as close as two lenses get at the same apertures. Heck, f/5.6, f/8, and f/11 on either lens is pretty much interchangeable with the other lens/aperture combos.

Now compare to the f/2.8L II. Is the f/2.8 version sharper? Yes. By a large amount? Not really. The difference is greatest at f/4 and still smaller then post processing choices.

The new 100-400 II certainly does better at 400mm then the 300 f/4L IS + 1.4x, but that's to be expected.

Here any prime should be noticably better than a general purpose zoom imho.

Based on what? The old adage that primes are always better then zooms hasn't been true in years thanks to modern lens design and manufacturing. The 100-400 II was a significant update and no primes blow it away now.

I therefore believe that a new version would be a significant improvement.

That would put it well above the f/2.8L II. I don't expect that to happen. There's not a lot of room to move here as all of the lenses we're discussing are quite sharp. If you can't produce hair splitting 20x30" prints with any of these lenses then something is wrong.

Its surely the weakest performing white prime lens in the current Canon line-up. No wonder given its age.

I have no idea where you got this impression of the 300 f/4L IS. It's a stellar performer. Even if we were to split line pairs and concluded that it was the "weakest" of the current line up, that would still make it better then 99% of lenses in existence. (Actually, the 400 f/5.6L is probably the "weakest" of the white lenses right now, and still an incredible performer.)

It will be interesting to see what the new 100-400 II does to sales of the 300 f/4L IS. But the 70-300L didn't knock demand or price down.

Lenses / Re: Why not uppdate just the IS
« on: January 17, 2015, 04:24:50 AM »
Why not just update the transmission in your 1991 Automobile?  There is more than just the IS that needs updating.  The design is good, but its 24 years old.  It does not have coatings on the lens elements to prevent reflections from the sensor, which reduces contrast.  It is also not all that sharp compared to newer lenses, and suffers from LOCA's or purple fringing.

Um...what?  ??? We are talking about the 300 f/4L IS USM, right?

One of the better lenses Canon makes. My only lens that's actually worth more today then when I bought it as market demand allowed Canon to push up the price.

It's as sharp as the much newer 70-300 f/4-5.6L IS USM and 100-400 II. It's not as sharp as the 300 f/2.8L IS USM II, but it's close enough that after post processing I doubt anyone could tell them apart. I'm struggling to think of lenses that are sharper, and among those I can think of (the f/2.8 version; Zeiss primes and Sigma ART primes) it's only by a small amount.

CA and flare are both very well controlled. I've never seen sensor reflections or lost contrast even in very harsh light. (The 300 f/4L IS is my go to surfing lens when shooting from a pier. Also my go to airshow lens. Those are about as harsh/high contrast as it gets.)

LOCA's are present but aren't bad at all. They are a fact of life with relatively fast lenses (f/4 is fast for 300mm), and there's only so much you can do to eliminate them.

Many sports shooters turn off IS in any event, since it does not help with moving subjects.

I don't know about anyone else, but IS helps me track the target, especially after a long day of shooting. If I'm hand holding a lens it's on. And it's a solid 2 stop / often 3 stop gain with this lens.

Canon could make it as sharp as its f/2.8 big brother and add 4 stop IS for an update. But it's darn near perfect as is.

EOS Bodies / Re: High Megapixel Camera Coming in 2015 [CR3]
« on: December 17, 2014, 10:56:36 AM »
I bet on it being a Canon sensor.  My guess is it will have the 5DIII AF System, and probably less than 4 fps.

I think I speak for everyone when I say that I hope it has no less then 14 stops of noise free DR...so the DR fights can stop  ;)

Only thing is, I do not care about any of the results shown in the samples in this thread at all. All of these images are falling apart - especially when they would be viewed large. And there is no more contrast or life left in them.

LOL! They are not "falling apart." I have a shadow exposure to compare my lifted version to, and though it's a little better, it's not substantially different.

Canon RAWs are fine to +2.5 and +3ev (depending on model). And the image which started yet another DRone debate could have easily been shot on a Canon.

Lenses / Re: Yet another DXO Interpretation Time video with Tony Northrup
« on: December 14, 2014, 02:53:38 AM »
To rebut, I'd first have to watch.  There are many things higher on my priority list than viewing Northrup's videos.  Watching paint dry and picking lint from my umbilicus are two things that come to mind...   :o

I don't know...it was fun to watch him back peddle on the 7D2. Both he and his wife tried it without any preconceived notions from DxO and loved it, the performance and the IQ.

Then when DxO said it was no better then a D300 (DxO b trollin' yo) he was in the uncomfortable position of reconciling his trust in DxO with his real world experience.

I can't help but wonder if the same thing happens in this video?

Lenses / Re: Review: Canon EOS 7D Mark II
« on: December 14, 2014, 01:14:27 AM »
The 7DII isn't intended for high ISO use. That's the 6D and 5DIII.

A crop sensor that's usable at ISO 16,000 is "not intended for high ISO use?"  ::)

Oh, and the 7DII isn't intended for low ISO, low noise landscape use either, that's intended for the 6D and 5D III.

It's absolutely fine for landscape use. 99% of the time you couldn't tell it apart from a D810 in a 24" print given good glass on both and low/mid ISOs.

Unless, of course, you habitually underexpose by 5ev.

Lenses / Re: Review: Canon EOS 7D Mark II
« on: December 14, 2014, 01:07:56 AM »
I was surprised how much the DR difference to the Nikon D7000 is, but then I remembered how well my old 7D can deal with harsh lighting conditions with the right PP technique. :D

DPReview's processing of the Canon file was awful. There are multiple samples in a thread on their forums that are much better and therefore closer to the D7000. Still not as good, but much closer.

Lenses / Re: Review: Canon EOS 7D Mark II
« on: December 14, 2014, 01:07:01 AM »
And even those may soon discover, that mirrorless cameras like Samsung NX1 are better suited for reach-limited, action-oriented captures.  :)

The Samsung cannot AF. In TheCameraStoreTV's review it blew half the frames in a dead simple tracking test (brisk walk/slow jog speed). In the Golden City Films review the two guys just tore the Samsung apart over AF. They said it was terrible at tracking at a dog park and even in the studio trying to get the 85mm to focus on the model's eye. The 7D II? Reliable as expected.

Absent a major firmware update, I don't think the NX1 is going to make any in roads into the Canon/Nikon sports and wildlife markets.

EOS Bodies - For Stills / Re: 6D vs 7D mark II
« on: December 09, 2014, 01:31:44 AM »
Not really.  None of the side-by-side comparisons I've seen between 7D and 7Dii are showing much improvement at all, and nothing has shouted out "much better", or even "after 5 years...". 

At low ISO the 7D2 is less 'gritty'. At high ISO I would say it's "much better", though obviously not as good as the 6D.

On the other hand, the difference between 7D + 10-22mm and 6D + 16-35mm is significant (I own both bodies and lenses).  I loved the combo with 7D, had lots of fun with it, but there is a big step up with the 6D.

That's because of the lens in question. A Tokina 11-16 or Sigma 8-16 would even that up.

Attached is a +3EV push (midtones) to +4EV push (shadows) from a Rebel T2i.

And that makes three pieces of hard evidence.

Again: Sony FF Exmor can stretch further. The shadow detail can be deeper...pitch black before processing...and still be recovered. But if you ETTR with Canon and can just start to see the detail to be recovered in the unprocessed version, you will typically be OK. There's a lot of room there, just not the same amount that's on Exmor FF.

you must be joking ...  +2.5 EV is a HUGE problem for Canon sensors, and most definitely for the 7D and all other APS-C sensors.

There are two shots in this thread proving you wrong. I realize some of the sliders are adaptive, but I reset Shadows to 0 and used just the Exposure slider to verify that my original settings were indeed equal to +2.5ev.

What do you imagine you can gain by stomping your feet and insisting X is false when faced with hard, reproducible evidence that X is true?

Also .. the image you have linked to is not even remotely comparable, and neither is the other one with the tree trunk. No sky at all in it, not to mention sun directly in the frame ... as in the linked picture in the starting post.

My image is directly comparable. In fact, looking at the unprocessed versions my shadow detail appears to be a little deeper then his. That he has the sun in the frame is immaterial given that he did not capture/preserve highlight detail. What patches of sky are visible in his shot are completely blown out.

I *strongly believe*

What you believe is irrelevant in the face of hard evidence no matter how strongly you feel about it.

Pages: [1] 2 3 ... 57